Causal mediation analysis of observational, population-based cancer survival data

Substantial socioeconomic inequalities in breast cancer survival persist in England, possibly due to more advanced tumour stage at diagnosis and differential access to treatment. However, both factors may be on the causal pathway between the socioeconomic level and cancer survival. Disentangling their respective contributions is therefore challenging when analysing such observational data.

The presentation will illustrate the importance of using causal inference methods and the need for testing key assumptions through sensitivity analyses. For example, we found that, for women diagnosed with breast cancer in England, one third of the short-term higher mortality and one tenth of the longer-term higher mortality experienced by most deprived patients was mediated by adverse stage distribution. And additional sensitivity analyses testing some of our study limitations showed that up to thirty per cent of the higher mortality in most deprived patients could be mediated by differential surgical treatment.

Our results suggest that, although effort for earlier diagnosis is important, this would reduce the cancer survival inequalities only by a third. Because of data limitations, role of differential surgical treatment may have been under-estimated.